“Angry Birds was Rovio’s 52nd game. ….Startups are hard, but they can also go from difficult to great incredibly quickly. You just need to survive long enough and keep going so you can create your 52nd game.” - Chris Dixon, “The Myth of the Overnight Success”
Rovio didn’t pivot. Rovio kept making games even in the face of bankruptcy in early 2009.
[This post is part of Startup Edition - see what other brilliant people in the startup world have to say about pivoting.]
But how do you know when it actually is time to pivot away from an idea? How do you know with confidence that you’re not onto something and your 52nd title is not around the corner? Indeed, the very same believing that keeps you going is very similar to the same believing that led you to take the initial leap in the first place. How do you ignore sunk costs?
It is not easy. But clearly we are all not Rovio. Capital, employee morale, and personal strength are all finite. We cannot all go on indefinitely.
In my experience, it boils down to a brief but complex question: have you disproved your core hypothesis?
My Story of Disproving Hypotheses
In June 2011, I started a website that recommends software to people based on personal needs and monetizes when people end up buying the software. You want a file sharing service that is the most affordable for 8 GB of storage and offers enterprise level security? TechPolish would recommend Box for you and earn about 8% of what you were worth to Box as a customer.
The core hypothesis: If people need a superior way to discover ideal software solutions, then people would use and share TechPolish to purchase software.
In the short 8 months I pursued TechPolish, it got some blog coverage, I made a number of feature improvements, many people told me they liked the product, and so on. Yet it broke down.
There were many places where the original hypothesis faltered:
If people need a superior way to discover ideal software solutions
I didn’t even make it this far. People didn’t need a superior way to discover ideal software solutions. Instead, some people may have kinda-sorta-wanted a better way, but that was it. After-the-fact user research showed me that people were perfectly fine Googling around for software vendors - my solution was not 10x better. Running a few Google Ads campaigns (and monitoring conversion rates to see if people had any interest in “Looking for an easy way to compare X, Y, and Z?”) could have helped me reach this conclusion faster.
then people would use and share TechPolish to purchase software.
People didn’t use the site that much based on time on site / average pages per visit. Moreover, my hope that people would geek out like I did and share these important sources of information on their social networks was flawed. Hardly any TechPolish pages led to Facebook shares / likes or Tweets or Linkedin Shares or Google +1’s. And the conversion to paid services was low, so people were not really using TechPolish to drive purchases either, or at least not in a first-degree manner.
Being Honest When Your Business Hypothesis is Disproven
When I realized all of this, I reached the important conclusion that the business’s economics would not lead to profitability. TechPolish did not have extremely unique or valuable content, so it did not rank well in search engines. TechPolish, for the same reasons, was not something people would share frequently. So it became unlikely that I could drive much organic traffic. And if the only avenue to pursue traffic was paid, I would be pursuing traffic through the same paid channels as companies who would be capturing 100%, not 8-15%, of the generated revenue. 8-15% is what you can earn through affiliate channels.
Profitability wouldn’t scale with paid marketing. So I began to move on from TechPolish but I have kept it running because the costs are so low and it is minorly profitable off of just organic traffic. I had crafted a small revenue stream, but with the core hypothesis disproved, it was time to move on.
Ultimately It Wasn’t a Pivot So Much as a Failure
Whether to pivot or to shut down is a choice often pondered in the same breath. You are moving away from your current pursuit, either way.
I plan to use this general approach of trying to disprove the core hypothesis as soon as possible when starting new businesses. In order to reach this point, you need to be intellectually honest and extremely aggressive about pursuing the truth.
When to Pivot and Not Shut Down
I chose to shut down instead of pivoting. Here are a couple factors that could have motivated me to pivot instead, if they were true:
- Traction in a subset of the product
For example, Yelp was originally an emailed recommendation service but noticed its users were using the service to share their own recommendations. The core hypothesis was disproved, but there was still a subset of the product that was getting traction. Groupon experienced the same when it pivoted from ThePoint.com.
What if I found people were using TechPolish as an enterprise market research platform vs. a consumer information portal? Maybe I would have pivoted to a SaaS offering for premium content and a robust interface for product comparisons and research.
- Shifting market trends
Suppose you were building a mobile security product around feature phones in 2007. You had product-market fit and customers. But now the iPhone came out. Smartphones grow in traction in 2007, more in 2008, even more in 2009. You wouldn’t succeed if you kept working toward feature phones. At some point in this period of time, you would pivot perhaps to mobile security for smartphones or tablets.
What if there was a faster explosion in SaaS offerings and for some reason the free content out there was falling behind? Maybe this also would have been a reason to become a premium content source and maintain the core of the business.
Have you pivoted a business? How do you even define “pivot”?
 Please note that disproven hypotheses and unproven hypotheses are different. Had I not felt like I had answered the key questions yet, the core hypothesis simply would have been unproven — not disproven. Then it would be more a question of remaining capital and passion to determine whether to double down or not and seek to finally disprove or prove the core hypothesis.